Substance Use and Delinquency

Abstract

Although there has been a decline in teen pregnancy in recent decades, reports have provided evidence that the decline is reversing. Adolescents involved in the child welfare system (CWS) are vulnerable to engage in high-risk behaviors including sexual behaviors, substance use, and delinquency. The literature on sexual risk behaviors in adolescents in the CWS has emphasized that such behaviors are associated with unplanned pregnancies; however, few studies have explored the co-occurrence of other risk behaviors and sexual behaviors. The purpose of this study was to determine the extent to which substance use and delinquent acts predict teen pregnancy of youths involved with the CWS. The sample consisted of 291 female adolescents in the National Survey of Child and Adolescent Well-Being study. Results indicated females who engage in substance use or delinquency at baseline were more likely to have been pregnant 36 months later than those who did not engage in such behaviors.

Keywords

delinquency substance use teen pregnancy

Adolescent pregnancy is an important social issue (Lamanna, 1999). Although there was a decline in teen pregnancy between 1991 and 2005, recent reports have provided evidence that the decline that lasted over a decade is now reversing (Farber, 2009). Between 2005 and 2007, there was a 5% rise in teen pregnancy rates among adolescents who were between 15 and 19 years old (Child Trends, 2010). Between 2007 and 2008, the United States experienced a 2% decrease; however, the rate of teen pregnancy within the United States is still one of the highest among developed countries (Alan Guttmacher Institute, 2011). In recent years, cumulatively more than 30% of women have become pregnant prior to the age of 20 (Kirby, 2007). Adolescents involved in the child welfare system (CWS) are more likely to engage in high-risk behaviors including sexual risk taking behaviors, substance use, and delinquency (Jones, 2010; Leslie et al., 2010). Sexual risk taking behaviors may include early sexual initiation, unprotected intercourse, engaging in sex with multiple partners, and engaging in intercourse while under the influence of drugs or alcohol (O’Donnell, Myint-U, Duran, & Stueve, 2010). The increased risk of engaging in high-risk sexual behaviors is often related to a history of maltreatment, unstable and dysfunctional family interactions, low socioeconomic status, as well as exposure to parental engagement in high-risk sexual behaviors or substance use (James, Montgomery, Leslie, & Zhang, 2009).

The existing literature on high-risk sexual behaviors in adolescents in the CWS has emphasized that sexual risk behaviors are often associated with unplanned or unwanted pregnancies (Hipwell, Keenan, Loeber, & Battista, 2010; Jones, 2010; O’Donnell et al., 2010); however, there is limited research that has explored the co-occurrence of other high-risk behaviors and sexual risk behaviors. Child Trends (2009) reported that youth involved with social service systems including the CWS are more likely to experience an unintended pregnancy. Using data from National Longitudinal Survey of Adolescent Health, results indicated that 17% of youth exposed to system services reported a teen pregnancy (Child Trends, 2009). The majority of research, however, has focused solely on youth placed in out-of-home care rather than all adolescents whose families have experienced an investigation by the welfare system (Leslie et al., 2010). Yet research demonstrates that adolescents involved with child welfare services, regardless of whether or not they were removed from the home, become teenage parents at substantially higher rates compared to adolescents among the general population (Vinnerljung, Franzen, & Danielsson, 2007). Therefore, the purpose of this study is to determine the extent to which substance use and delinquent acts among adolescent females predicted teen pregnancy of youths involved with the CWS.

Teen Pregnancy

In recent years, the age of initiation into sexual activity for female adolescents has progressively decreased. Early initiation of sex, prior to the age of 14 years, is commonly related to lower usage of protection and higher rates of teen pregnancy (Hipwell et al., 2010). It is important to note that there are biological characteristics that may influence early initiation of sexual intercourse. Factors such as estrogen and androgen levels, age of first menstrual cycle, and signs of early development have all been related to early initiation (Farber, 2009).

Research has demonstrated the importance of family to youths’ decisions to delay sexual activity (Aspy et al., 2010), and adolescents who are at the highest risk of early or unwanted pregnancy live in family environments and possess personal characteristics that lead them to participate in high-risk behaviors such as substance use or delinquent behaviors (Vinnerljung et al., 2007). Decreasing the current rates of teen pregnancy requires a better understanding of the risk behaviors that may precede or co-occur with high-risk sexual activity (Hipwell et al., 2010). Adolescents involved with the CWS have often experienced abuse, neglect, or general dysfunction. Exposure to these factors makes this population more likely to engage in high-risk behaviors such as substance use, acts of delinquency, and risky sexual behaviors (Leslie et al., 2010).

Youth that are more likely to participate in high-risk sexual behaviors (including unprotected sex) often have lower expectations and poorer performance in school, engage in problem behaviors (acting out, antisocial, or other delinquent behaviors), are easily encouraged by peers to participate in problem behaviors, have weak bonds with their parents, and reside in lower income neighborhoods (Dryfoos, 1990).

Unintended teen pregnancy is associated with several negative outcomes. Infants born to teen mothers tend to be born prematurely, weigh less at birth, and experience increased rates of mortality. Furthermore, children are more likely to have poorer academic and behavioral outcomes relative to children born to mothers in their 20s and 30s (Child Trends, 2010). Examples of poor outcomes include lower math and reading scores; worse communication, language, and social skills; and lower physical and emotional well-being. Children of teen parents are more likely to engage in intercourse at younger ages and become teen mothers themselves. Teen pregnancy also has a negative impact on the teen mothers. For example, teen mothers are less likely to complete high school or go to college. Teen mothers are more likely to depend on financial assistance such as welfare (National Campaign to Prevent Teen Pregnancy, 2002). Furthermore, teens who have one child may be more likely to have additional children in their teen years (Child Trends, 2010). For example, Sims and Luster (2002) reported that of adolescent mothers who intended to reduce the likelihood of a second pregnancy and who received family support services, approximately 60% reported a repeat pregnancy. In addition, teen mothers are less likely to marry, and those who do marry are more susceptible to divorce (Child Trends, 2010).

Although extant research has demonstrated important factors associated with teenage pregnancy (Heflinger, Simpkins, & Combs-Orme, 2000), studies examining the relationships among specific risk factors for adolescent females involved with CWS are scarce (James et al., 2009). Studying these risks is particularly important considering the elevated risks for negative outcomes for adolescents involved with the CWS.

Theoretical Background

Hirschi’s (1969) social control theory guided this study. Social control theory posits that four factors of social bonds restrain individuals from engaging in deviant behaviors (i.e., substance use, acts of delinquency, and risky sexual behaviors investigated in the present study): attachment to others, commitment, involvement in structured activities, and conventional beliefs (values). Attachment relates to an adolescent’s ability to establish bonds with significant others. Commitment involves an adolescent’s ability to have goals or look positively toward the future. Involvement constitutes an adolescent’s ability to participate in activities that allow him or her to feel successful or goal oriented. Lastly, conventional beliefs relates to an adolescent’s acceptance of social rules (Wiatrowski, Griswold, & Roberts, 1981). According to social control theory, individuals who create an attachment to their environment, such as their family, friends, or school, will be less likely to deviate from age-appropriate societal norms. If an individual can create a positive attachment to at least one aspect of his or her environment, he or she will be less likely to engage in deviant behavior (Farber, 2009). In contrast, youth with low levels of social control and high delinquent peer associations, for example, are at an even greater risk of committing delinquent acts (Lin, Cochran, & Mieczkowski, 2011). Notably, adolescents involved in the CWS are more likely to engage in high-risk behaviors including sexual risk taking behaviors, substance use, and delinquency (Jones, 2010; Leslie et al., 2010). A recent study designed to identify risk factors associated with female adolescent delinquency, for example, found that lower levels of family trust and caring were predictive of higher rates of delinquency (Cernkovich, Lanctot, & Giordano, 2008). Furthermore, ineffective parenting also has been found to predict adolescent delinquency (Simons, Chao, Conger, & Elder, 2001). When applying social control theory to adolescents involved with the CWS, theoretically these youth may have a diminished sense of social control because of weakened social bonds (e.g., family and peers). As a result, increases in problem behaviors may be expected. Because of the relationship between substance use and delinquency, and the engagement of high-risk sexual behaviors for adolescent females, these variables were the focus of the present study.

Substance Use and Teen Pregnancy

Early initiation of alcohol use, as a factor alone, is an issue of concern to society at large because of its lifelong implications. Alcohol use at early ages has served as an indicator of personal vulnerabilities as well as a pathway to other substance use and engagement in other high-risk behaviors such as violence, theft, and truancy (O’Donnell et al., 2010). Early initiation in substance use has been related to several unfavorable outcomes associated with high-risk sexual behaviors including increased risk of contracting a sexually transmitted infection as well as unplanned pregnancy (O’Donnell et al., 2010). Specifically for youth involved with the CWS, use of both “gateway” drugs, including alcohol, cigarettes, and marijuana, and “other” drugs, including heroin and cocaine, were associated with significantly increased odds of engaging in sexual risk behaviors (James et al., 2009).

Several studies have shown that substance use and teen pregnancy co-occur (Cavazos-Rehg et al., 2010; Leslie et al., 2010; O’Donell et al., 2010; Taylor-Seehafer & Rew, 2000); however, most research has focused on substance use as it relates to high-risk sexual behaviors in general. These articles cite unwanted teen pregnancy as an outcome of the engagement in high-risk behavior and also include other factors such as number of sexual partners, failure to use protection from sexually transmitted infections, and substance use during sexual intercourse (O’Donnell et al., 2010). Dryfoos (1990) reported that substance abuse is related to early sexual activity and that teen pregnancy is related to early sexual activity and heavy substance use.

Delinquency and Teen Pregnancy

Research has demonstrated an association between delinquency and the early onset of sexual activity (e.g., Dryfoos, 1990; Hockaday, Crase, Shelley, & Stockdale, 2000). For example, a study conducted by Auslander and colleagues (2002) found a relationship between externalizing behaviors and engagement in high-risk sexual behaviors. Similarly, Pederson and Mastekaasa (2010) found a strong association between conduct disorder symptoms and pregnancy among a sample of Norwegian girls. In a prospective study comparing a group of pregnant adolescents to a group of adolescents who were not pregnant, pregnant adolescents were significantly more likely to have engaged in higher risk taking behaviors including alcohol use and delinquent behaviors (Hockaday et al., 2000). Few studies, however, have specifically examined these factors with youth involved with the CWS. Given the increased risk for sexual risk taking behaviors, substance use, and delinquency among youth involved with the CWS (Jones, 2010; Leslie et al., 2010), gaining a better understanding of these relationships within this at-risk population seems particularly important.

The present study involved secondary data analyses of the National Survey of Child and Adolescent Well-Being (NSCAW; 2002). The NSCAW is a national study of the well-being of more than 5,000 children involved in the CWS across the country. This study examined two potential risk behaviors related to teen pregnancy—substance use and delinquency—for adolescents involved with the CWS. The goal of this study was to explore the extent to which substance use and delinquent acts predicted teen pregnancy 36 months later.

Method

Survey Design and Sample

The target population of NSCAW was children aged 0 to 14 years in the United States for whom an investigation of child abuse or neglect had been opened within a 15-month period between October 1999 and December 2000. A two-stage cluster sampling design was used for the study (NSCAW, 2002). First, the United States was divided into nine sampling strata and within each stratum primary sampling units were formed. Ultimately, 92 sampling units were selected that represented distinct geographic areas across the United States. The final sampling units represented 96 counties in 36 states. In the participating counties, children that fit the inclusion criteria were randomly selected from the population of the children so that the same number of children was selected from each unit, regardless of the size of the unit (NSCAW, 2002). Of the total participants, 2,769 were female. Of these females, 683 participants were over the age of 11, and 346 had responded to items regarding substance use, delinquency, and pregnancy. As we were interested in determining the extent to which substance use and delinquent behaviors could predict pregnancy 36 months later, we excluded females who reported experiencing a pregnancy prior to or at Wave 1 (n = 46). The final study sample included 291 female adolescents who had responded to the question regarding whether or not they had been pregnant in their life during the 36-month follow-up and who were not pregnant during Wave 1 of the study. For the current study, analyses were conducted on this subset of female adolescents. Because of the longitudinal nature of the NSCAW study, we were able to examine participants’ report of engagement in at-risk behaviors of delinquency and substance use at Wave 1 (baseline) and the outcome variable of teen pregnancy at Wave 4 (36 months later). At Wave 4, the mean age was 15.74 with a minimum of 13 and maximum of 18 (SD = 1.19). The sample was racially diverse: 52.2% Caucasian, 32.3% African American, 4.1% American Indian, 1.7% Asian/Hawaiian/Pacific Islander, and 9.7% “other.” Furthermore, 14.4% of the sample identified their ethnicity as Hispanic.

Measures

Pregnancy

Pregnancy was a dichotomous variable measured through an item assessing if the adolescent had ever been pregnant during the course of their lifetime (0 = no or 1 = yes). This study used this assessment at Wave 1 and Wave 4 of the study.

Substance use

Substance use was measured through 14 items gathered through a modified version of the Drug Free School Community Act Outcome Study Questions (U.S. Department of Education: Office of the Under Secretary, as cited in NSCAW, 2002) completed by the youth at Wave 1. The measure included 14 items that assessed how frequently the youth engaged in the use of substances over the course of their lifetime as well as over the past 30 days (1 = 1 day to 6 = 20 or more days). Questions assessing substance use over lifetime asked youth to self-report how many days they used seven substances including alcohol, cigarettes, chewing tobacco, marijuana, inhalants, hard drugs, and prescription drugs. A follow-up question then assessed how many days they had used the substance in the past 30 days. The measure is divided into two subscales: substance use over the youth’s lifetime and substance use during the past 30 days. For the purposes of this study, the reliability coefficients were α = .92 and α = .81, respectively.

Delinquent behaviors

For the purpose of this study, delinquency was defined through self-report of criminal involvement (theft, physically attacking another individual, fraud, being disorderly in public) as well as engagement in status offenses (running away from home, truancy) by female adolescents in the CWS. A modified version of the Self Report of Delinquency (Elliott & Ageton, 1980, as cited in NSCAW, 2002), measuring specific delinquent acts and the frequency of these acts, was completed by youths at Wave 1. The modification was developed from the Self Report of Delinquency originally used in the National Longitudinal Survey of Youth, which was a nationally representative study of 12,686 male and female participants between the ages of 14 and 22 years. The measure included 72 questions. These 72 questions measured two aspects of delinquency: one scale that asked youth about committing acts of delinquency (yes or no) and a second scale that assessed how frequently youth committed such acts (1 = once to 5 = 5 or more times). The questions fall into one of nine categories: felony assault, felony theft, robbery, fraud, minor theft, illegal services, damage to property, public disorder, and status offenses. Questions first assessed if youth committed acts of delinquency. This scale assessed overall engagement in delinquent behaviors. Questions then assessed how frequently youth engaged in delinquent acts such as attacking another individual, stealing, selling drugs, damaging property, being unruly, or skipping school. For the purpose of the study, a final raw score on the 72 items was by computed by the NSCAW. Within the NSCAW data set, the reported internal consistency was high (α = .98) for the total sample. For the purpose of the current study, the reliability coefficient for the scale was α = .88.

Analyses

Initial analyses were conducted to determine the prevalence of the dependent variable within the sample at Wave 4 (if the adolescent had ever been pregnant; n = 82). Of the 291 youth, 82 (nearly 28% of the sample) reported that they had been pregnant at least once in their life during Wave 4 of the study. We then conducted an ANOVA to determine if the variable of race was associated with the likelihood of a female adolescent ever being pregnant. The results indicated that race did not significantly affect the likelihood of experiencing a pregnancy, F(1, 289) = .175, p > .05. For the purpose of this study, type of maltreatment and placement type were not controlled for because the sample included all adolescents who had an open case with CWS, regardless of whether or not allegations of maltreatment were substantiated. Table 1 provides the means and standard deviations for the study variables for the whole sample. A series of logistic regressions were conducted to determine if youths’ engagement in substance use over their lifetime, and substance use over the past 30 days, delinquent acts, and the frequency in which they engaged in delinquent acts significantly predicted teen pregnancy 36 months later. Including highly correlated (i.e., more than .70) variables in the same model could result in statistical error (Freedman, Pee, & Midthune, 1992). As can be expected given the measures of the present study assessed identical behaviors across different durations, substance use over their lifetime and substance use over the past 30 days were highly correlated, r(289) = .85, p < .001, and thus were explored separately. Similarly, total delinquency and the frequency of engagement in delinquent behavior, r(179) = .92, p < .01, also were explored separately because of the high correlation among variables.

Table 1.

Reports of Female Adolescents in the Child Welfare System, Youth Problem Behaviors and Substance Use: Descriptive Statistics (N = 291).

Variables	M	SD	Range
Total delinquent acts	4.5	4.07	0-20
Total delinquency score	7.63	12.64	1-80
Felony assault	.38	1.15	0-7
Status offenses	1.74	2.83	0-15
Substance use: Lifetime	12.20	8.44	0-43
Substance use: Within 30 days	4.08	5.42	0-25

Results

Substance Use

Logistic regression was conducted to estimate if substance use over lifetime significantly predicted teen pregnancy. Substance use over the youth’s lifetime significantly predicted teen pregnancy (χ² = 4.435, df = 1, n = 291, p < .05). Next, a logistic regression was conducted to estimate if substance use during the past 30 days significantly predicted teen pregnancy. Substance use during the past 30 days significantly predicted teen pregnancy (χ² = 5.991, df = 1, n = 291, p < .05). Substantively, adolescents who reported more frequent substance use over their lifetimes, as well as during the past 30 days, were more likely to report having been pregnant by Wave 4 of the study (see Table 2).

Table 2.

Summary of Logistic Regression Analysis for Variables Predicting Teen Pregnancy.

Predictor	B	SE B	e^B	p
Overall engagement in delinquent behaviors	.025	.010	1.025	.015
Frequency of engagement in delinquent acts	.070	.038	1.073	.064
Felony assault	.271	.111	1.312	.015
Status offenses	.141	.046	1.151	.002
Substance use over lifetime	.032	.015	1.032	.035
Marijuana	.144	.054	1.115	.008
Alcohol	.112	.060	1.119	.063
Prescription drugs	.163	.082	1.117	.048
Substance use over past 30 days	.057	.023	1.058	.014

Note. e^B = exponentiated.

After examining substance use over their lifetime and substance use within the past 30 days, we examined if female adolescents who had reported being pregnant were more likely to engage in the use of a particular substance over the course of their lifetime. Logistic regression was conducted, and the results indicated that females who used marijuana were more likely to report a pregnancy (χ² = 6.862, df = 1, n = 291, p < .05). Use of alcohol and prescription drugs approached significance (see Table 2).

Delinquent Behaviors

Logistic regression analyses were conducted to estimate if engagement in delinquent behaviors at baseline significantly predicted teen pregnancy 36 months later. Delinquent behaviors were assessed by overall engagement in delinquent behaviors (total delinquency) and the frequency (number of acts) of delinquent behaviors. Overall engagement in delinquent behaviors significantly predicted teen pregnancy (χ² = 6.049, df = 1, n = 291, p < .05; see Table 2). The number of times the youth engaged in delinquent behaviors (total delinquent acts) did not significantly predict teen pregnancy. Substantively, adolescents who reported any engagement in delinquent behaviors were more likely to have been pregnant by Wave 4; however, the frequency that they engaged in the delinquent behaviors did not affect their likelihood of being pregnant by Wave 4 (see Table 2).

After examining the overall delinquency and delinquent acts scores, we examined each of the nine categories (felony assault, felony theft, robbery, fraud, minor theft, illegal services, damage to property, public disorder, and status offenses) to determine if there was a specific type of delinquent behavior in which the female adolescents who became pregnant were most likely to participate. Logistic regression was conducted, and the results indicated that the sample was most likely to commit the following acts: felony assault (χ² = 6.031, df = 1, n = 291, p < .05) and status offenses (χ² = 9.316, df = 1, n = 291, p < .05).

Discussion

Research has demonstrated that adolescents involved with CWS become teenage parents at substantially higher rates compared with adolescents among the general population (Vinnerljung et al., 2007). Given the potential for a myriad of negative outcomes associated with teenage pregnancy, it is important to understand risk factors specifically for this vulnerable population.

By exploring substance use and delinquency as predictors of teen pregnancy, the results of the current study contribute to the literature in several ways. First, prior studies have explored the relationship between substance use, delinquency, and high-risk sexual behaviors; however, few studies have focused on these two high-risk behaviors as they relate specifically to teen pregnancy, particularly for this high-risk population. This study adds to the literature by not only demonstrating a relationship among drug use, delinquency, and teen pregnancy, but also by suggesting a potential correlation between these high-risk behaviors and pregnancy. Additionally, the majority of research has focused solely on youth placed in out-of-home care rather than all adolescents whose families have experienced investigations by the welfare system. Yet past research has demonstrated the elevated risks of teen pregnancy for adolescent females who experience negative family situations (e.g., Lamanna, 1999). Findings from this study provide support for focusing on all adolescents involved with CWS, not just youths removed from adverse family conditions. Lastly, a rise in pregnancy rates from prior decades among adolescents aged 15 to 17 years requires a reinvestigation of behaviors that may be associated with teen pregnancy in order to inform prevention programs targeted for this population. Results of this study indicated that substance use and delinquency may be antecedent behaviors associated with a later increased likelihood of teen pregnancy.

Female adolescents involved in the CWS who reported engaging in substance use at baseline were more likely to become pregnant 36 months later compared with peers who did not engage in this high-risk behavior. Furthermore, the use of marijuana specifically was also associated with a greater likelihood of having been pregnant by Wave 4. Previous research that has demonstrated a relationship between substance use and early sexual activity (Cavazos -Rehg et al., 2010; Leslie et al., 2010; O’Donell et al., 2010; Taylor-Seehafer & Rew, 2000) support these findings.

The results also suggest that female adolescents involved in the CWS who reported engaging in delinquent behaviors at baseline were more likely to become pregnant 36 months later compared with their peers who did not engage in such behaviors, and past research has demonstrated a similar trend (e.g., Auslander et al., 2002; Dryfoos, 1990). Interestingly, findings of the present study suggest that the frequency in which adolescents engaged in delinquent acts did not affect the likelihood of becoming pregnant. The results reveal that an adolescent who self-reported minimal levels of engagement in delinquent behaviors (once in the past 6 months) was equally likely to report having been pregnant by Wave 4 as an adolescent who reported high levels (five or more times in the past 6 months). Lastly, female adolescents who engaged in delinquent behaviors such as felony assault (attacking an individual with the intent to hurt or kill them) or status offenses (truancy, running away from home) at baseline were more likely to report being pregnant by Wave 4 of the study. Although the analyses in this study were correlational, by using Wave 1 data as predictors of Wave 4 behaviors, results may begin to indicate important pathways to teen pregnancy through engagement in high-risk behaviors.

Although earlier studies have placed emphasis on youths removed from the home (Dworsky, 2009; James et al., 2009), findings from this study suggest that adolescents involved with CWS (not just those removed from the home) are at risk to experience negative outcomes such as teen pregnancy as a result of the engagement in substance use and delinquent behaviors. Lamanna (1999) asks, “Can we devise policies and family supports for actors whose behavior responds to limited options” (p. 210). Lamanna’s question, formulated based on research with adolescents in the general population, certainly seems applicable to adolescents involved with the CWS, who arguably may have even fewer options than peers not involved with CWS. It may be important to provide primary prevention programs for females involved with CWS and to evaluate the extent to which participation in these efforts prevents future adolescent pregnancy.

There is an ongoing debate about how and if teen pregnancy prevention programs should be publicly or privately funded; however, it is noted that programs should not be geared only toward the prevention of teen pregnancy but enhancing youth development as a whole. At-risk youth may require more aggressive efforts because of the increased likelihood of teen pregnancy in this population. Efforts need to address specific issues such as truancy, mental health issues, and substance use issues (Farber, 2009).

When considering Hirschi’s (1969) social control theory as it relates to the findings of this study, it is important to consider factors that might contribute to an adolescent having a higher likelihood of engaging in deviant behaviors. A lack of attachment to others, commitment, involvement in structured activities, and values contribute to adolescents’ increased likelihood of engaging in high-risk behaviors. Prevention strategies for adolescents in the CWS that promote a sense of connection and community could be a means of deterring the engagement in delinquent behaviors. Consistent with social control theory, research on the influence of the family on adolescent sexual initiation (Aspy et al., 2010; Vinnerljung et al., 2007) suggests that family- or group-centered interventions may help adolescents foster a sense of attachment, commitment, or involvement.

Prevention programs addressing sexual antecedents tend to focus on things such as attitudes, beliefs, and knowledge about sex; behavioral strategies for refusing to have sex or unprotected sex; and access to contraception (Farber, 2009). However, if adolescent substance use and delinquency are risk factors associated with an increased risk of teen pregnancy, perhaps these behaviors may also be important behavioral antecedents that warrant attention in pregnancy prevention efforts. Furthermore, prevention efforts need to be developed in a manner that is sensitive to the stage of development, values, and characteristics of the community as there is no one type of program that will serve this entire population (Farber, 2009). Although much more research is needed to understand the nature of this relationship, focusing interventions toward youths who commit more serious infractions may leave adolescents with less serious acts of delinquency without potentially beneficial services.

Limitations and Future Research

There are limitations in the current study that are important to consider. First, the current analyses were conducted with a subset of the original sample. Although the NSCAW data set includes more than 5,000 individuals, this study examined a very specific subsample: adolescent females who reported being pregnant during the 36-month follow-up of the study but who had not been pregnant at Wave 1. Although the study only included a subset of the NSCAW population, using Wave 1 data to predict Wave 4 outcomes specifically for adolescents involved with the CWS had not been established in previous studies using these data. Next, although we examined a variety of delinquent behaviors, as well as used multiple substances as predictors, delinquency and substance use were the only two high-risk behaviors analyzed for the study. Although past research has connected these two variables to high-risk sexual behaviors, the researchers deemed it important to make a connection between these two types of high-risk behaviors and teen pregnancy specifically; however, important variables not considered in the present study, such as socioeconomic status, and whether or not allegations of maltreatment were substantiated, may also meaningfully affect the results. Future researchers could consider these additional characteristics.

The NSCAW data set relies on the self-report of the participants for the measures of interest in this study. The accuracy of self-report data rests on a participant’s willingness and ability to disclose information. Rosenthal and Curiel (2006) found that children’s self-report of problem behaviors in the NSCAW sample were higher than caregiver reports. Therefore, using adolescent report in the present study may have resulted in higher levels of problem behaviors than if reported by the caregiver. Despite this finding by Rosenthal and Curiel, adolescents may be less likely to share about engagement in behaviors such as violence, theft, robbery, and substance use. Thus, youth report may be a more true reflection of behaviors within this sample. Furthermore, this study used a measure of substance use assessing behaviors within the past 6 months; however, it is impossible to know how soon the teen became pregnant after the measures were completed. Although the study demonstrated a relationship between substance use and pregnancy, future studies that could provide more information on the frequency of substance use prior to conception may contribute to the research. In general, the measurement of these variables in future studies is an important consideration and area for improvement.

Conclusion

The purpose of this study was to determine the extent to which substance use and delinquent acts among adolescent females predicted teen pregnancy of youths involved with the CWS. Our focus was on adolescents involved in the CWS, not just adolescents placed in out-of-home care. The results indicated that females who engaged in substance use or delinquency at baseline were more likely to have been pregnant 36 months later than those who did not engage in such behaviors. These findings may suggest a need to establish programs tailored to at-risk youth, their families, and communities and address issues such as delinquency and substance use in order to best serve this population.

Footnotes

Authors’ Note

This article includes data from the National Survey on Child Adolescent Well-being, which was developed under contract with Administration on Children, Youth, and Families, U.S. Department of Health and Human Services (ACYF/DHHS). The data were provided by the National Data Archive on Child Abuse and Neglect. The information and opinions expressed herein reflect solely the position of the authors. Nothing herein should be construed to indicate the support or endorsement of its content by ACYF/DHHS.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

The Alan Guttmacher Institute. (2011). Facts on American teens’ sexual and reproductive health. Retrieved from http://www.guttmacher.org/pubs/FB-ATSRH.pdf

Aspy

Vesely

Tolma

Oman

Rodine

Marshall

Fluhr

(2009). Youth assets and delayed coitarche across developmental age groups. Journal of Early Adolescence, 30, 277-304.

Auslander

McMillen

Elze

Thompson

Johnson-Reid

Stiffman

(2002). Mental health problems and sexual abuse among adolescent in foster care: Relationship to HIV risk and behavior intentions. AIDS and Behavior, 6, 351-359.

Cavazos-Rehg

Spitznagel

Krauss

Shootman

Bucholz

Cottler

Bierut

(2010). Understand adolescent parenthood from a multisystemic perspective. Journal of Adolescent Health, 46, 525-531.

Cernkovich

S. A.

Lanctot

Giordano

(2008). Predicting adolescent and adult antisocial behavior among adjudicated delinquent females. Crime and Delinquency, 54, 3-33.

Child Trends 2009 Vulnerable youth: A closer look at reproductive health. Retrieved from http://www.thenationalcampaign.org/resources/pdf/SS/SS42_VulnerableYouth.pdf

Child Trends. (2010). Teen births. Retrieved from http://www.childtrendsdatabank.org/?q=node/52

Dryfoos

J.G.

(1990). Adolescents at risk: Prevalence and prevention. New York, NY: Oxford University Press.

Dworsky

(2009). Preventing pregnancy among youth in foster care. Congressional Roundtable with Senator Mary Landrieu, Russell Senate Office. Retrieved from http://www.chapinhall.org/sites/default/files/DworskyFosterPregnancy-7-16-09.pdf

10.

Farber

N. B.

(2009). Adolescent pregnancy; Policy and prevention services (2nd ed.). New York, NY: Springer.

11.

Freedman

L. S.

Pee

Midthune

D. N.

(1992). The problem of underestimating the residual error variance in forward stepwise regression. Statistician, 41, 405-412.

12.

Heflinger

C. A.

Simpkins

C. G.

Combs-Orme

(2000). Using the CBCL to determine the clinical status of children in state custody. Children and Youth Services Review, 22, 55-73.

13.

Hipwell

A. E.

Keenan

Loeber

Battista

(2010). Early predictors of sexually intimate behaviors in an urban sample of young girls. Developmental Psychology, 46, 366-378.

14.

Hirschi

(1969). Causes of delinquency. Berkeley: University of California Press.

15.

Hockaday

Crase

S. J.

Shelley

M. C.

Stockdale

D. F.

(2000). A prospective study of adolescent pregnancy. Journal of Adolescence, 23, 423-438.

16.

James

Montgomery

S. B.

Leslie

L. K.

Zhang

(2009). Sexual risk behaviors among youth in the child welfare system. Children and Youth Services Review, 31, 990-1000.

17.

Jones

T. C.

(2010). It drives us to do it: Pregnant adolescents identify drivers for sexual risk Tasking. Issues in Comprehensive Pediatric Nursing, 33, 82-100.

18.

Kirby

(2007). Emerging answers 2007: Research findings on programs to reduce teen pregnancy and sexually transmitted diseases. Washington, DC: National Campaign to Prevent Teen Pregnancy.

19.

Lamanna

M. A.

(1999). Living the postmodern dream: Adolescent women’s discourse on relationships, sexuality, and reproduction. Journal of Family Issues, 20, 181-217.

20.

Leslie

L. K.

James

Monn

Kauten

M. C.

Zhang

Aarons

(2010). Health-risk behaviors in young adolescents in the child welfare system. Journal of Adolescent Health, 47, 26-34.

21.

Lin

W. H.

Cochran

J. K.

Mieczkowski

(2011). Direct and vicarious violent victimization and juvenile delinquency: An application of general strain theory. Sociological Inquiry, 81, 195-222.

22.

National Campaign to Prevent Teen Pregnancy. (2002). Not just another single issue: Teen pregnancy prevention’s link to other critical social issues. Washington, DC: Author.

23.

National Survey of Child and Adolescent Well-Being. (2002). National Survey of Child and Adolescent Well-Being (NSCAW), Introduction to the Wave 1 general and restricted use releases. Retrieved from http://www.ndacan.cornell.edu/NDACAN/Datasets/UserGuidePDFs/092_Intro_to_NSCAW_Wave_1.pdf

24.

O’Donnell

Myint-U

Duran

Stueve

(2010). Especially for daughters: Parent education to address alcohol and sex-related risk taking among urban young adolescent girls. Health Promotion Practice, 11, 70S-78S.

25.

Pederson

Mastekaasa

(2010). Conduct disorder symptoms and subsequent pregnancy, child-birth and abortion: A population-based longitudinal study of adolescents. Journal of Adolescence, 34, 1025-1033.

26.

Rosenthal

J. A.

Curiel

H. F.

(2006). Modeling behavioral problems of children in the child welfare system: Caregiver, youth, and teacher perceptions. Children and Youth Services Review, 28, 1391-1408.

27.

Simons

R. L.

Chao

Conger

R. D.

Elder

G. H.

(2001). Quality of parenting as mediator of the effect of childhood defiance on adolescent friendship choices and delinquency: A growth curve analysis. Journal of Marriage and Family, 63, 63-79.

28.

Sims

Luster

(2002). Factors related to early subsequent pregnancies and second births among adolescent mothers in a family support program. Journal of Family Issues, 23, 1006-1031.

29.

Taylor-Seehafer

Rew

(2000). Risky sexual behavior among adolescent women. Journal for Specialist in Pediatric Nursing, 5, 15-25.

30.

Vinnerljung

Franzen

Danielsson

(2007). Teenage parenthood among child welfare clients: A Swedish national cohort study of prevalence and odds. Journal of Adolescence, 30, 97-116.

31.

Wiatrowski

Griswold

Roberts

(1981). Social control theory and delinquency. American Sociological Review, 46, 525-541.